In many applications it is desirable to characterize quasi-periodic signals. Quasi-periodic signals may be produced from a variety of sources, one principle source being piston engines or vehicles driven by piston engines. One way of characterizing (i.e. extracting features) the source or vehicle is to find the amplitude of the base frequency and the various harmonics making up the quasi-perodic signal. Once the quasi-periodic signal is characterized, the quasi-periodic signal can be compared to known signals, through standard pattern classification techniques, to identify the source of the quasi-periodic signal.
Generally, when sources of quasi-periodic signals, such as piston engines and the like, vary in speed, traveling conditions, etc. the base frequency (piston firing rate) and harmonics vary so that the characteristic pattern of the source varies. In the prior art, discrete-time spectrum analyzers are utilized in conjunction with a fixed clock frequency to analyze quasi-perodic signals. The problem with this type of analysis is that when conditions of the source are altered (speed, terrain, etc.) the base frequency and harmonics of the quasi-periodic signal are altered and the analysis can no longer be compared to a predetermined pattern.